Bug 2023977

Summary:	Duplicate SMBIOS handles when creating large VMs
Product:	Red Hat Enterprise Linux 9	Reporter:	Eduardo Habkost <ehabkost>
Component:	qemu-kvm	Assignee:	Ani Sinha <ani>
qemu-kvm sub component:	Devices	QA Contact:	Yiqian Wei <yiwei>
Status:	CLOSED ERRATA	Docs Contact:
Severity:	unspecified
Priority:	medium	CC:	ailan, ani, chayang, coli, imammedo, jinzhao, juzhang, mrezanin, nilal, virt-maint, xuwei, yiwei
Version:	unspecified	Keywords:	TestOnly, Triaged
Target Milestone:	rc
Target Release:	9.1
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	qemu-kvm-7.0.0-1.el9	Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2022-11-15 09:53:29 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:
Bug Depends On:	2064757
Bug Blocks:	1788991
Deadline:	2022-05-23

Description Eduardo Habkost 2021-11-16 21:48:46 UTC

When creating VMs with a large amount of RAM, the SMBIOS tables created by QEMU may contain duplicate handle IDs.

Problem diagnosed at:
https://bugzilla.redhat.com/show_bug.cgi?id=1904267#c51

Copying from original report:

(In reply to Eduardo Habkost)
> (In reply to Brian Payton)
> > ASSERT_EFI_ERROR (Status = Already started)
> > ASSERT
> > /builddir/build/BUILD/edk2-ca407c7246bf/OvmfPkg/SmbiosPlatformDxe/
> > SmbiosPlatformDxe.c(125): !EFI_ERROR (Status)
> 
> This one might be an unexpected duplicate SMBIOS handle in the tables
> generated by QEMU.  I will double-check the SMBIOS table generation code to
> be sure.

Yes, this is where the problem comes from:


static void smbios_build_type_17_table(unsigned instance, uint64_t size)
{
    SMBIOS_BUILD_TABLE_PRE(17, 0x1100 + instance, true); /* required */

...
static void smbios_build_type_19_table(unsigned instance,
                                       uint64_t start, uint64_t size)
{
    SMBIOS_BUILD_TABLE_PRE(19, 0x1300 + instance, true); /* required */


Using the current handle assignment code, we can have only up to 512 DIMM slots before their type 17 SMBIOS handles conflict with the type 19 table handles.  QEMU has DIMM sizes hardcoded to 16GB, which means an 8TB guest will hit the limit.

Comment 1 John Ferlan 2021-11-23 19:02:15 UTC

Assigning to Amnon for reassignment, NB: module is maintained by mst.

Comment 2 Ani Sinha 2022-02-01 10:37:17 UTC

I am going to take a stab at this. Could not assign the bug to myself unfortunately (I am not redhat :()

Comment 3 Ani Sinha 2022-02-01 11:13:05 UTC

The problem is in this code in smbios_get_tables()

       dimm_cnt = QEMU_ALIGN_UP(current_machine->ram_size, MAX_DIMM_SZ) / MAX_DIMM_SZ;                                                                                         
                                                                                                                                                                                
        smbios_build_type_16_table(dimm_cnt);                                                                                                                                   
                                                                                                                                                                                
        for (i = 0; i < dimm_cnt; i++) {                                                                                                                                        
            smbios_build_type_17_table(i, GET_DIMM_SZ);                                                                                                                         
        }                                                                                                                                                                       
                                                                                                                                                                                
        for (i = 0; i < mem_array_size; i++) {                                                                                                                                  
            smbios_build_type_19_table(i, mem_array[i].address,                                                                                                                 
                                       mem_array[i].length);                                                                                                                    
        }                                                                                                


Clearly after 512 iterations of the first loop, the two tables will collide as 0x1100 + 512 = 0x1300.

Comment 4 Ani Sinha 2022-02-02 04:39:28 UTC

Sent a patch in the mailing list: https://lists.gnu.org/archive/html/qemu-devel/2022-02/msg00358.html

Comment 5 Ani Sinha 2022-02-02 13:40:57 UTC

@ailan@ailan please re-assign this ticket to me.

Comment 6 Amnon Ilan 2022-02-02 16:06:48 UTC

(In reply to Ani Sinha from comment #5)
> @ailan@ailan please re-assign this ticket to me.

Done. Thank you for taking it!

Comment 9 Nitesh Narayan Lal 2022-02-03 13:29:45 UTC

Moving this to rhel9, if the fix is going to come as part of the rebase, then we can mark this as TestOnly.
For rhel8, we will create a clone if required.

Comment 12 Ani Sinha 2022-03-09 10:57:13 UTC

A fix has been pushed upstream: https://git.qemu.org/?p=qemu.git;a=commit;h=a379d4550c4a949bb1ca51bc5272d68b4a619b18 . QE, please verify.

Comment 13 Nitesh Narayan Lal 2022-03-09 16:48:40 UTC

(In reply to Ani Sinha from comment #12)
> A fix has been pushed upstream:
> https://git.qemu.org/?p=qemu.git;a=commit;
> h=a379d4550c4a949bb1ca51bc5272d68b4a619b18 . QE, please verify.

Thanks, Ani, for getting the fix upstream.
I am moving this BZ back to ASSIGNED. Once the patch is merged upstream please add the commit ID in the comment and then move the BZ to POST.

Since the patches will automatically come via reabse there is no action required after that. QE will test it as the images are available.
Igor has already provided a repo with the fix to the reporter, and hopefully, we will hear something soon.

Comment 14 Ani Sinha 2022-03-10 06:29:03 UTC

@nilal the patch has already been merged upstream. That is why I updated the ticket. 

Commit ID: a379d4550c4a949bb1ca51bc5272d68b4a619b18 as I have mentioned above.

Comment 15 Nitesh Narayan Lal 2022-03-10 11:35:02 UTC

(In reply to Ani Sinha from comment #14)
> @nilal the patch has already been merged upstream. That is why I
> updated the ticket. 
> 
> Commit ID: a379d4550c4a949bb1ca51bc5272d68b4a619b18 as I have mentioned
> above.

Got it, thanks for the correction.

Comment 21 errata-xmlrpc 2022-11-15 09:53:29 UTC

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Moderate: qemu-kvm security, bug fix, and enhancement update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2022:7967